Package $

Source Code of $.CrawljaxRunner

package ${package};

import com.crawljax.browser.EmbeddedBrowser.BrowserType;
import com.crawljax.condition.UrlCondition;
import com.crawljax.core.CrawljaxController;
import com.crawljax.core.configuration.CrawlSpecification;
import com.crawljax.core.configuration.CrawljaxConfiguration;
import com.crawljax.core.configuration.InputSpecification;
import com.crawljax.core.configuration.ThreadConfiguration;

/**
* Use the sample plugin in combination with Crawljax.
*
*/
public class CrawljaxRunner {

  private static final String URL = "http://www.google.com";
  private static final int MAX_DEPTH = 2;
  private static final int MAX_NUMBER_STATES = 8;
 
  public static void main(String[] args) {

    try {
      // configure the crawling engine
      CrawljaxConfiguration config = getCrawljaxConfiguration();

      // add your plugin
      config.addPlugin(new SamplePlugin());

      // initilize and run Crawljax
      CrawljaxController crawljax = new CrawljaxController(config);
      crawljax.run();

    } catch (Exception e) {
      e.printStackTrace();
    }
  }
 
  private static CrawljaxConfiguration getCrawljaxConfiguration() {
    CrawljaxConfiguration config = new CrawljaxConfiguration();
    config.setCrawlSpecification(getCrawlSpecification());
    config.setThreadConfiguration(getThreadConfiguration());
    config.setBrowser(BrowserType.firefox);
    return config;
  }

  private static ThreadConfiguration getThreadConfiguration() {
    ThreadConfiguration tc = new ThreadConfiguration();
    tc.setBrowserBooting(true);
    tc.setNumberBrowsers(1);
    tc.setNumberThreads(1);
    return tc;
  }

  private static CrawlSpecification getCrawlSpecification() {
    CrawlSpecification crawler = new CrawlSpecification(URL);
    crawler.setRandomInputInForms(false);
    // click these elements

    crawler.click("a");
    crawler.click("button");

    // except these
    crawler.dontClick("a").underXPath("//DIV[@id='guser']");
    crawler.dontClick("a").withText("Language Tools");

    crawler.setInputSpecification(getInputSpecification());

    // limit the crawling scope
    crawler.setMaximumStates(MAX_NUMBER_STATES);
    crawler.setDepth(MAX_DEPTH);

    return crawler;
  }

  private static InputSpecification getInputSpecification() {
    InputSpecification input = new InputSpecification();
    input.field("gbqfq").setValue("Crawljax");
    return input;
  }
}
TOP

Related Classes of $.CrawljaxRunner

TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.